Saturday, March 28, 2009

Beware of so-called Linux proponents

Every now and then you stumble across a blog that is run by a so-called Linux enthusiast. Some of them claim to have been using both Linux and Windows for years, so they have balanced and objective view on the advantages and disadvantages of both systems. But when you start to look a little closer, you will see that they spread the SOFUD. Some have a real gift for writing and are so credible that you see no need to investigate their claims.

Mike Dailey, a Cisco engineer, is one of these bloggers. Although he has been blogging for only three months, you will find some very interesting articles here, like "Why Enterprise Adoption of Linux Is Slow", "Why You Shouldn’t Use Linux" and "The Need for Linux, Microsoft and Open Source".

His love for Linux is well known. He tells us over and over again he is a real Linux proponent and wishes Linux and FOSS all the best in the future, which he has well drawn out.
We need Microsoft leading the way because of their market share in so many software technologies, but we need them to stop trying to take over the world and begin to show technical leadership of the industry. Microsoft could easily take Linux and Open Source under their wing, to help this alternative mature and to help guide them down an avenue that is best for everyone.

After j00p34 published his "10+1 things to tell your boss why you should migrate to Linux" he could help himself and had to give poor j00p34 some advise.
Here the author supports the well-known “Linux is more secure” argument. There is no real basis for the opinionated argument as stated by the author, with no quantifiable facts or data to support the notion that Linux is superior in terms of security.

I was eager to find any links in his blog substantiating this, but unfortunately I found none. He continues with:
Any Linux distro, given to an inept admin with lacking security skills, will be far less secure than an out-of-the-box Windows server platform.

Again, I expected him to come up with any significant figures. Nothing. Now have read quite some German magazines and I happen to remember several articles on the subject when Vista came out. Verdict: OS/X was best out of the box, Linux came next and both Vista and XP were last.

His argument "the user is to blame" is a well known piece of SOFUD. It goes like this: even if your house is a fortress, it won't do you any good if you leave the door open. So since people tend to leave the door open, it won't matter whether you use paper or steel doors. Thus, paper doors are sufficient.

Another piece of SOFUD is that there are no figures on this subject. Wrong again. You just have to know where to look. This report from IBM shows that Apache is far more secure than IIS. This report from Google confirms it. It seems to be a common characteristic of FOSS, since this shows that Firefox users are less at risk than IE users. You say this isn't about Linux, Mike? What about this one. Couldn't you come up with something, Mike, or are you just too lazy to research your story properly? Sorry, Mike, please continue..
If you present the “stability” argument to management you must be prepared to present uptime reports and outage root-cause analysis data to back up your argument. If you are experiencing severe outages in Windows servers in the data center your cause likely resides with the skills of your administrative staff, not your server operating system.

Mikey, Mikey, Mikey.. Don't keep on using the same old tricks. First, don't blame it on the admin, and second, don't tell me the data is not available. The whole internet is monitored by Netcraft, showing that your blog is run by Apache and Linux. Fortunately, Nicholas Petreley did the analysis for me, so I don't have to waste any more time debunking your unfounded post:
The average uptime of the Windows web servers that run Microsoft’s own web site ( is roughly 59 days. The maximum uptime for Windows Server 2003 at the same site is 111 days, and the minimum is 5 days. Compare this to (a sample site that runs on Linux), which has had both an average and maximum uptime of 348 days.

Ok, I could go on and waste some more time on TCO or code quality - and if you're not nice to me, I might even do that - but I think, I will leave it at this and just refer to a professional that put up the following testimonial on his site:
Migrated a multitude of Windows NT/2000 systems to Red Hat Linux to lower TCO and enhance system stability and performance. Oracle 9i RAC, Checkpoint firewall, IBM Websphere Commerce are examples of systems migrated to Red Hat Linux.

But hey, this is the resume of Mike Dailey himself! Surprise, surprise. I know, Mike, you didn't like me finding that out. That's why you deleted my comments. Next time, be more careful, will you.

Update: Mike Dailey has written a followup on his article called "The Death of the Linux Debate: A Eulogy". He makes some good points there, which I have addressed in my own followup. I have a good idea of what the concerns of management are involving the application of Open Source, since I have to deal with them professionally in what may be the most FOSS unfriendly country in the world: The Netherlands.

Update: Mike and me might be closer to each other than we thought. Please read his excellent comment here.

Update: I had to change this article slightly, because some so-called FOSS supporters don't seem to know when enough is enough.

Update: Mike Dailey has truly given a worthy closing to this debate. Although I cannot undo this post, I cannot honestly maintain that Mike Dailey fits this profile.


Anonymous said...

Its been a while but I think I heard somewhere that Linux uptime counters reset to zero* after 348 days. So unless my memory is faulty or they fixed that the real figure could be much much higher.

* Some BSDs have the same problem others don't.

Debianero said...

Great article.

Xackery said...

Well, what really can you expect when you read a blog or article than hearing some one's opinion leaning one way or another? I think the real word of advice you should be giving than discrediting another article is simply say remember these are all opinions. I recommend reading plenty from both sides of the gate, and figure out for yourself the solution to whatever problems you may have that each operating system can condone.

Some times, both can solve your problem, so then you really should make a pros/cons list for your specific situation, and see what your boss says. But hey, that's my opinion, just like this is your opinion about the other guy's opinion which was about another guy's opinion. Gotta love opinions.

I am here said...

Why do people post links to security reports from 2-5 years old? 2 years in software development is a lifetime and those reports are no longer valid.

Anonymous said...

Haha this is from his own resume.

"Migrated a multitude of Windows NT/2000 systems to Red Hat Linux to lower TCO and enhance system stability and performance."

So he says there is no proof of Linux being more stable or having smaller TCO?

And it surely does have in many cases. French police Gendarmerie Nationale declared 70% savings this year on IT budget.

Anonymous said...

You, my poor misguided friend, are a retard.

Anonymous said...

Some people are just too scared to try something new. They also usually have some inertia as well, and they better stick to it vs. seriously investigating an alternative, especially if it is a better one in so many categories.

This explains why their minds will find and collect some facts or just out-right assumptions to support their "stickiness" to their status quo.

We should not blame Michael for his inability to change his status quo quickly. Instead we should encourage his free spirit to try new things. Mike, find time, make it work. You won't regret it. ;-)

On the different note, just installed ubuntu 9.04 netbook remix beta on my Asus Eee 901. Boy, this thing flies! Wow! Out-of-the box it recognized everything on my Eee. I am lovin' it. ;-) It should show Mike that FOSS is even better from a business perspective too, because it allows to quickly react to market changes and deliver a basically new version of OS that meets a new category of devices, such as netbooks or MIDs.

Thanks to all FOSS developers and supporters!

Anonymous said...

It's a well known 32 bit counter overflow,

However, it doesn't cause the computer to reboot. So, if your box has been up nearly two years and your upgtime shows only 90 days then add 497 to it to compute the real number of days it has been running.

woodbastwick road said...

Dear Mike,

Due to the pervasiveness inherent to FOSS, you have quite an uphill struggle ahead, take a deep breath and tackle your slippery slope head on. I hope FOSS continues to supply you with useful tools to carry out your livelihood, the way it does with countless others the world over, bringing with them the means to improve your understanding of them, something that is somewhat more difficult to do with closed source software.

M3EEKS said...

I'm really tired of this kind of thing.
There are black-hats hiding among us! Conspiracies everywhere. It's war, you know?
I exclusively use Linux - Mepis & LinuxMint, and you know what? I have yet to encounter more than a few FOSS applications & Linux Distros I'd be willing (if I had to) pay for. Id est: the quality for an AVERAGE end-user is subpar & never gets beyond beta quality. There's no one in control of direction & quality. Developer's think forcing end-users to edit config files is happy, time-wasting perfection of design. Elegance & ease of use are for suckers. That's the nature of FOSS. You get what you pay for.
amaroK 1.4 I'd gladly pay $20 for. Evolution, if I used an email client. Open Office? It's on par w/ MS Office 2003 in quality & function.$30? OSes? I don't know.
It's not FUD that prevents widespread adoption of Linux & FOSS; it's Linux & FOSS that does so. We've come a long way, but till a complete, elegant, fully functional, truly production-quality os shows up (Android?) that the general population just has to have, we're still servers & hobbyists. Non-integrated elements slapped together do not a great OS make. Netbook OS? Windows XP. We missed that boat after a short ride back to shore (Thanks gOS for chasing away a whole bunch of users!).
There's no way in Hell my Mom would be using Mint on her laptop if I (geek) weren't around. FOSS, in general? I'm not really sure.
These are a few things to ponder, instead of looking for the enemy among us & lying to ourselves & the world about the FOSS & Linux Utopias.

Anonymous said...


Editing config files by hand? Dude, 1998 called, and they want their argument back. Linux is about as "out of the box and onto the 'net" as Mac OS is these days. Arguing otherwise only shows that you either haven't bothered to try Linux lately (which means you're unqualified to comment about it) or you have a vested interest in keeping the Abusive Monopolist of Redmond afloat.

No matter, though; commoditization is a difficult force to overcome. It may be taking a long time but it's happening.

DWL said...

Presumably this kind of blog is paid for by a sizeable back-hander in a pub car park somewhere?

Anonymous said...

Uptime like 384 days might be a sign of unprofessionalism as patches should be installed at least quarterly.

Low uptime of windows server might be partially explainable by higher freq. of patching required by Windows.

Anonymous said...


Why should you need a restart for a patch?

unless you're talking about a kernel patch.

Anonymous said...

If it's a windows patch, you probably do need to restart.

As the above poster said, unless it's a kernel patch (or maybe a libc patch), then you don't need to do a reboot on *nix. Given that most exploitation happens fairly high up in the stack (ie. at the server process or web application level), critical kernel/libc updates aren't common.

sims said...

First of all, nice article. It's good to hear this kind of thing. FUD and authors of FUD should be openly discredited and humored.

Secondly, anybody who complains about having to edit a file is a user and should not be allowed to admin any system. Well, maybe a fax machine. Actually, I take that back. These days you can ssh into the fax. OK, so go admin your pencil. Arg...

Go back and study or bury your head in the TV. I am sick and tired of the noise of the masses of clueless people. I now understand projects that do not want to use forums and choose mailing lists or IRC, etc. Web 2.0 sux. Let's quickly move on to web 3.0 so that the users are completely out of touch, where they belong, in this point and click world they live in.

M3EEKS said...

I use Linux soley. I don't own Windows or Mac OS.
True, it is NOT 1998 - one doesn't get stuck editing config files as much, but it's still required, unless you're really lucky, or have have a simple set-up. True, progress has been extremely fast (think back 5 years ago). It's much better, but it's still not where it needs to be. MythTV - easy to config, right? No one ever HAS to edit xorg.conf to get something working? I did w/ my last install. I shouldn't have had to, since there are controls in the DE that are supposed to adjust things. Samba? Sometimes it works right away, depending on the distro, but still, w/ in the last year I had to hack at that stupid config file several times on different installs. Mom (special example, I'll admit) wouldn't have her Linux Mint latop printing thru her Windows PC, if I hadn't known what to do. Code is better w/ open source, but often times design is lacking & distros are just 'off-the shelf' components slapped together. I'm not bashing anything here, I'm just stating what I think are real reasons Linux & FOSS aren't on 90% of PCs in the world. We need to stop looking for conspiracies & always being on the defensive. There are real reasons why FOSS & Linux aren't dominating. Price is on our side. It's not all fear of change; it's not because MS is just out to get us. If OSX were some free linux distro,or their DE were a free download, do you think the world would be using Windows a all? Get what I'm saying?